Integrated circuit production relies on the use of photolithographic processes to define the active elements and interconnecting structures on microelectronic devices. Until recently, g-line (436 nm) and I-line (365 nm) wavelengths of light have been used for the bulk of microlithographic applications. However, in order to achieve smaller dimensions of resolution, the wavelength of light used for microlithography in semiconductor manufacturing has been reduced into the deep UV regions of 256 nm and 193 nm. The problem with using deep UV wavelengths is that resists used at the higher wavelengths were too absorbent and insensitive. Thus, in order to utilize deep UV light wavelengths, new resist materials with low optical absorption and enhanced sensitivities were needed.
Chemically amplified resist materials have recently been developed through the use of acid-labile polymers in order to meet the above-mentioned criteria. They have shown great promise in increasing resolution. However, chemically amplified resist systems have many shortcomings. One problem is standing wave effects, which occur when monochromatic deep UV light is reflected off the surface of a reflective substrate during exposure. The formation of standing waves in the resist reduces resolution and causes linewidth variations. For example, standing waves in a positive resist have a tendency to result in a foot at the resist/substrate interface reducing the resolution of the resist.
In addition, chemically amplified resist profiles and resolution may change due to substrate poisoning. Particularly, this effect occurs when the substrate has a nitride layer. It is believed that the N--H bond in the nitride film deactivates the acid at the nitride/resist interface. For a positive resist, this results in an insoluble area, and either resist scumming, or a foot at the resist/substrate interface.
Furthermore, lithographic aspect ratios require the chemically amplified resist layer be thin, e.g., about 0.5 .mu.m, to print sub 0.18 .mu.m features. This in turn requires the resist to have excellent plasma etch resistance such that resist image features can be transferred down into the underlying substrate. However, in order to decrease absorbance of the chemically amplified resist, aromatic groups, such as those in novolaks had to be removed, which in turn decreased the etch resistance.
Utilizing an underlayer or undercoat film that is placed on the substrate before the chemical amplified film is applied can reduce the above-mentioned problems. The undercoat absorbs most of the deep UV light attenuating standing wave effects. In addition, the undercoat prevents deactivation of the acid catalyst at the resist/substrate interface. Furthermore, the undercoat layer can contain some aromatic groups to provide etch resistance.
In the typical bilayer resist process, the undercoat layer is applied on the substrate. The chemically amplified resist is then applied on the undercoat layer, exposed to deep UV light and developed to form images in the chemically amplified resist topcoat. The bilayer resist system is then placed in an oxygen plasma etch environment to etch the undercoat in the areas where the chemically amplified resist has been removed by the development. The chemically amplified resist in a bilayer system typically contains silicon and is thus able to withstand oxygen plasma etching. After the bottom layer is etched, the resist system can be used for subsequent processing such as non-oxygen plasma etch chemistry which removes the underlying substrate.
Even though the undercoat attenuates standing waves and substrate poisoning, it poses other problems. First, some undercoat layers are soluble to the chemical amplified resist solvent component. If there is intermixing between the top and undercoat layers, the resolution and sensitivity of the top resist layer will be detrimentally affected.
In addition, if there is a large difference in the index of refraction between the chemical amplified resist and the undercoat layer, light will reflect off the undercoat layer causing standing wave effects in the resist. Thus, the index of refraction between the two layers must be matched to minimize reflectivity effects.
Another problem with undercoating layers is that they are sometimes too absorbent because of incorporation of aromatic groups. Some semiconductor manufacturing deep UV exposure tools utilize the same wavelength of light to both expose the resist and to align the exposure mask to the layer below the resist. If the undercoat layer is too absorbent, the reflected light needed for alignment is too attenuated to be useful. However, if the undercoat layer is not absorbent enough, standing waves may occur. A formulator must balance these competing objectives.
In addition, some undercoats have very poor plasma etch resistance to plasma chemistry. The etch resistance of the undercoat should be comparable to the etch rate of novolak resins in order to be commercially viable.
Furthermore, some undercoat layers require UV exposure in order to form cross-links before the radiation sensitive resist topcoat layer can be applied. The problem with UV cross-linking undercoat layers is that they require long exposure times to form sufficient cross-links. The long exposure times severely constrain throughput and add to the cost of producing integrated circuits. The UV tools also do not provide uniform exposure so that some areas of the undercoat layer may be cross-linked more than other areas of the undercoat layer. In addition, UV cross-linking exposure tools are very expensive and are not included in most resist coating tools because of expense and space limitations.
Some undercoat layers are cross-linked by heating. However, the problem with these undercoat layers is that they require high curing temperatures and long curing times before the top layer can be applied. In order to be commercially useful, undercoat layers should be curable at temperatures below 250.degree. C. and for a time less than 180 seconds. After curing, the undercoat should have a high glass transition temperature to withstand subsequent high temperature processing.
Therefore, it is an object of the present invention to provide a thermally curable polymer composition that is useful for an undercoat layer in deep UV lithography. Another object of the present invention is to provide an undercoat layer, which is cured at temperatures less than about 250.degree. C. and for a time less than about 3 minutes. It is a further object of this invention to provide an undercoat layer which is insoluble to the top resist's solvent system, minimizes reflectivity effects, and has an etch rate comparable to novolaks.
Other and further objects, advantages and features of the present invention shall become apparent as described below.